| Author |
Message |
|---|
|
|
|
Welcome to the Equinox Challenge
People spending a lot of time on the internet will recognize the meme "Brace yourself, winter is coming!", as will readers and watchers of the Game of Thrones books and tv series. For those on the southern hemisphere that's what this Equinox challenge is all about. First the fall, then winter. For us on the Northern hemisphere though it's all about leaving winter behind us, nature starting to grow and showing us all its springtime beauty once again. Following that, we are very much looking forward to finding many new primes during this Equinox Challenge.
A 5 day Challenge is being offered on PrimeGrid's Proth Prime Search (LLR) application.
To participate in the Challenge, please select only the Proth Prime Search (LLR) project in your PrimeGrid preferences section. The challenge will begin 17 March 2013 11:02 UTC and end 22 March 2013 11:02 UTC. Application builds are available for Linux , Windows and MacIntel 32 bit and 64 bit. CPU's with AVX capabilities will be significantly faster than the ones without, as this instruction set allows for more computing power.
ATTENTION: The primality program LLR is CPU intensive; so, it is vital to have a stable system with good cooling. It does not tolerate "even the slightest of errors." Please see this post for more details on how you can "stress test" your computer. WU's will take ~15 minutes on fast/newer computers and 2+ hours on slower/older computers. If your computer is highly overclocked, please consider "stress testing" it. Sieving is an excellent alternative for computers that are not able to LLR. :)
Please, please, please make sure your machines are up to the task.
[UPDATE] Highly overclocked Sandy Bridge computers running the latest llrAVX application will see faster times.
Time zone converter:
The World Clock - Time Zone Converter
NOTE: The countdown clock on the front page uses the host computer time. Therefore, if your computer time is off, so will the countdown clock. For precise timing, use the UTC Time in the data section to the left of the countdown clock.
Scoring Information
Scores will be kept for individuals and teams. Only work units issued AFTER 17 March 2013 18:00 UTC and received BEFORE 22 March 2013 18:00 UTC will be considered for credit. We will use the "prime score" method which is based on the n value (k*b^n-1) to score the challenge. The only difference is that the primary and double checker of a WU will receive the same score.
Therefore, each completed WU will earn a unique score based on its n value. The higher the n, the higher the score. This is different than BOINC cobblestones! A quorum of 2 is NOT needed to award Challenge score - i.e. no double checker. Therefore, each returned result will earn a Challenge score. Please note that if the result is eventually declared invalid, the score will be removed.
For details on how the score is calculated, please see this thread.
At the Conclusion of the Challenge
We kindly ask users "moving on" to ABORT their WU's instead of DETACHING, RESETTING, or PAUSING.
ABORTING WU's allows them to be recycled immediately; thus a much faster "clean up" to the end of an LLR Challenge. DETACHING, RESETTING, and PAUSING WU's causes them to remain in limbo until they EXPIRE. Therefore, we must wait until WU's expire to send them out to be completed.
Please consider either completing what's in the queue or ABORTING them. Thank you. :)
About the Proth Prime Search
The Proth Prime Search is done in collaboration with the Proth Search project. This search looks for primes in the form of k*2^n+1. With the condition 2^n > k, these are often called Proth primes. This project also has the added bonus of possibly finding factors of "classical" Fermat numbers or Generalized Fermat numbers. As this requires PrimeFormGW (PFGW) (a primality-testing program), once PrimeGrid finds a prime, it is then tested on PrimeGrid's servers for divisibility.
Proth Search only searches for k<1200. PrimeGrid created an extension to that which includes all candidates 1200<k<10000 for n<5M. It is this extension which we call PPSE that the Challenge will be on.
Initially, PrimeGrid's PPS project's goal was to double check all previous work up to n=500K for odd k<1200 and to fill in any gaps that were missed. We have accomplished that now and have increased it to n=1.3M. PG's LLRNet searched up to n=200,000 and found several missed primes in previously searched ranges. Although primes that small did not make it into the Top 5000 Primes database, the work was still important as it may have led to new factors for "classical" Fermat numbers or Generalized Fermat numbers. While there are many GFN factors, currently there are only 293 "classical" Fermat number factors known. Current primes found in PPS definitely make it into the Top 5000 Primes database.
For more information about "Proth" primes, please visit these links:
About Proth Search
The Proth Search project was established in 1998 by Ray Ballinger and Wilfrid Keller to coordinate a distributed effort to find Proth primes (primes of the form k*2^n+1) for k < 300. Ray was interested in finding primes while Wilfrid was interested in finding divisors of Fermat number. Since that time it has expanded to include k < 1200. Mark Rodenkirch (aka rogue) has been helping Ray keep the website up to date for the past few years.
Early in 2008, PrimeGrid and Proth Search teamed up to provide a software managed distributed effort to the search. Although it might appear that PrimeGrid is duplicating some of the Proth Search effort by re-doing some ranges, few ranges on Proth Search were ever double-checked. This has resulted in PrimeGrid finding primes that were missed by previous searchers. By the end of 2008, all new primes found by PrimeGrid were eligible for inclusion in Chris Caldwell's Prime Pages Top 5000. Sometime in 2009, over 90% of the tests handed out by PrimeGrid were numbers that have never been tested.
PrimeGrid intends to continue the search indefinitely for Proth primes.
What is LLR?
The Lucas-Lehmer-Riesel (LLR) test is a primality test for numbers of the form N = k*2^n − 1, with 2^n > k. Also, LLR is a program developed by Jean Penne that can run the LLR-tests. It includes the Proth test to perform +1 tests and PRP to test non base 2 numbers. See also:
(Edouard Lucas: 1842-1891, Derrick H. Lehmer: 1905-1991, Hans Riesel: born 1929).
____________
PrimeGrid Challenge Overall standings --- Last update: From Pi to Paddy (2016)
|
|
|
|
|
17 March 18:00 till 22 March 18:00
Shouldn't the time be 11:02 instead of 18:00? The countdown clock agrees with me.
____________
My lucky numbers are 121*2^4553899-1 and 3756801695685*2^666669±1
My movie https://vimeo.com/manage/videos/502242 |
|
|
 Dave 
Send message
Joined: 13 Feb 12
Posts: 3171
ID: 130544
Credit: 2,233,444,322
RAC: 550,199
                           
|
|
Yes if it's 18:00 I can fit in 2 more WRs... If it's 11:02 it's too tight for 2. |
|
|
|
|
Yes if it's 18:00 I can fit in 2 more WRs... If it's 11:02 it's too tight for 2.
@Dave - I think the challenge will be a full 5 days and the question that Pooh Bear is asking is the time it starts. The series announcement and the front page state that this challenge is from 11:02 - 11:02 (ie just like the normal 18:00 - 18:00) for 5 days but the info thread below shows it starts at 18:00 and goes through 18:00.
I think you will still be able to get your 2 wu's in if I'm reading your post correctly and you can start when the challenge starts.
@mad cow - Just to give you a little grief since you and your herd have run off to a bigger pasture :)
Cheers Rick
____________
@AggieThePew
 |
|
|
|
|
Yes if it's 18:00 I can fit in 2 more WRs... If it's 11:02 it's too tight for 2.
@Dave - I think the challenge will be a full 5 days and the question that Pooh Bear is asking is the time it starts. The series announcement and the front page state that this challenge is from 11:02 - 11:02 (ie just like the normal 18:00 - 18:00) for 5 days but the info thread below shows it starts at 18:00 and goes through 18:00.
I think you will still be able to get your 2 wu's in if I'm reading your post correctly and you can start when the challenge starts.
@mad cow - Just to give you a little grief since you and your herd have run off to a bigger pasture :)
Cheers Rick
Yarrr, no problem. This cow is super busy and copy/pasted just that little too much from the last PPS challenge. Fixed now though. Thanks for pointing it out all.
____________
PrimeGrid Challenge Overall standings --- Last update: From Pi to Paddy (2016)
|
|
|
|
|
|
Hooray for 11:02!! Starting at 21:02 local is much better than the normal 0400 local. I will not have to set the alarm clock for this one.
happy challenge to all.
____________
Member team AUSTRALIA
My lucky number is 9291*2^1085585+1 |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3171
ID: 130544
Credit: 2,233,444,322
RAC: 550,199
|
Yes if it's 18:00 I can fit in 2 more WRs... If it's 11:02 it's too tight for 2.
@Dave - I think the challenge will be a full 5 days and the question that Pooh Bear is asking is the time it starts. The series announcement and the front page state that this challenge is from 11:02 - 11:02 (ie just like the normal 18:00 - 18:00) for 5 days but the info thread below shows it starts at 18:00 and goes through 18:00.
I think you will still be able to get your 2 wu's in if I'm reading your post correctly and you can start when the challenge starts.
Not at 99 hours each I won't with the 2nd card only finishing now ;). It's 8 days away i.e 2×96 hours (4 units in total)...
Regards |
|
|
|
|
|
Will only CPU shrubs be counted or are the Beta Cuda shrubs available too?
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
                           
|
Will only CPU shrubs be counted or are the Beta Cuda shrubs available too?
If anyone is actually able to run the CUDA app, is willing to run it, and it returns a valid result, yes, it will count.
There are no llrCUDA apps available for download from the server, so the only way to run it is with app_info.
I do not recommend checking the llrCUDA beta checkbox; if you're going to use app_info I recommend using the regular CPU selection and overriding that with app_info. It seems that selecting the beta checkbox has a nasty habit of making the scheduler not send you any work, or send you the wrong work.
IMHO this is a waste of perfectly good GPU resources; it's not very fast and uses a lot of CPU time. Might as well just run on the CPU cores.
While we're talking about this...
(pause)
EDIT: There's no more llrCUDA box to check. You can still use llrCUDA by using app_info if you so desire.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Are we going to continue working on this 100<k<500 range for the challenge so that we have longer WUs (thus saving the server) or are we going to switch to higher k values that have lower n values than the current range being tested?
____________
 |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
Are we going to continue working on this 100<k<500 range for the challenge so that we have longer WUs (thus saving the server) or are we going to switch to higher k values that have lower n values than the current range being tested?
I'm not aware of any plan to change PPS ranges.
____________
My lucky number is 75898524288+1 |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3171
ID: 130544
Credit: 2,233,444,322
RAC: 550,199
|
EDIT: There's no more llrCUDA box to check. You can still use llrCUDA by using app_info if you so desire.
In which case bring on the GFN WRs :D!
We need more CUDA.
Think I'm going to parade the streets with a banner (might look a bit strange though)... |
|
|
|
|
Are we going to continue working on this 100<k<500 range for the challenge so that we have longer WUs (thus saving the server) or are we going to switch to higher k values that have lower n values than the current range being tested?
I'm not aware of any plan to change PPS ranges.
100<k<2000 is now being sent out, so it does appear that the range is changing slightly for the challenge.
____________
|
|
|
Ken_g6Volunteer developer
Send message
Joined: 4 Jul 06
Posts: 938
ID: 3110
Credit: 258,807,858
RAC: 92,400
                     
|
|
I tried the very latest version of llrCUDA on a 15*2^N-1 number the other day. The result was truly unimpressive - barely faster than a single Core 2 core on a GTX 460, and nowhere near an AVX-enabled CPU.
I wonder why CUDA is so much better for GFN than LLR?
____________
|
|
|
|
|
|
A WARNING FOR EVERYONE THINKING OF PARTICIPATING IN THE CHALLENGE!!!
If you have PPS sieve GPU tasks enabled while participating in this challenge then you most likely will lose days of crunching unless you are babysitting your PC.
Due to a bug in the boinc server software that has recently gotten much worse the server WILL send you pps sieve cpu tasks if at some point your PC requests CPU and GPU tasks at the same time.
The best way to avoid this during the challenge if you don't have the time to babysit is to run GFN GPU tasks. |
|
|
|
|
|
j.sheridan
PPS ( LLR ) - the best tasks ever for hungry CPUs
and GPUs ( PPS ( sieve ) I dont stop it, I have no problem with BoinC Server.
Everyone now thinks that makes the Boinc server problems.
The PC can also be spezifiziwer, depending from OS
If you mean GFN only,
GFN is runing / on my OS |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3171
ID: 130544
Credit: 2,233,444,322
RAC: 550,199
|
|
Sounds a nice plan defind1 - sounds like you'll have a consistent & high RAC as well :). |
|
|
|
|
j.sheridan
PPS ( LLR ) - the best tasks ever for hungry CPUs
and GPUs ( PPS ( sieve ) I dont stop it, I have no problem with BoinC Server.
Everyone now thinks that makes the Boinc server problems.
The PC can also be spezifiziwer, depending from OS
If you mean GFN only,
GFN is runing / on my OS
Me 2.
I have PPS Sv and LLR running with no problems. I am beginning to wonder why me!!
If I do get the wrong ones I will just send them back (abort) as I would with anything else that was delivered from somewhere when I had not ordered it.
____________
Member team AUSTRALIA
My lucky number is 9291*2^1085585+1 |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
Are we going to continue working on this 100<k<500 range for the challenge so that we have longer WUs (thus saving the server) or are we going to switch to higher k values that have lower n values than the current range being tested?
There's a mix of work in there, so you'll get both PPS (low K, high N) and PPSE (high K, low N) work. We just added more PPSE work, which is shorter workunits, to give everyone a better chance of finding primes during the challenge.
I'm not concerned about the server handling the load. I do expect there to be some temporary "no work" messages at the beginning of the challenge, simply because of the way the BOINC system works. It pauses for no less than 1 second once the buffer is full -- so if enough requests come in during that second, it's possible to empty the buffer before the feeder wakes up again. Any more requests that come in before the feeder fires up after it's 1-second nap will get that error and have to ask again.
Ironically, this is MORE of a problem now that the server is a lot faster. It's a lot easier to empty the entire buffer in under a second with a fast server than a slow server.
____________
My lucky number is 75898524288+1 |
|
|
|
|
To participate in the Challenge, please select only the Proth Prime Search (LLR) project in your PrimeGrid preferences section. The challenge will begin 17 March 2013 11:02 UTC and end 22 March 2013 11:02 UTC.
...
Only work units issued AFTER 17 March 2013 18:00 UTC and received BEFORE 22 March 2013 18:00 UTC will be considered for credit.
Question:
What exactly does "only work units *issued* AFTER ..." means? If I change now the project properties to only get that task, but keep it running my current PPS Sieve queue (around 9h), will I not be credited/scored even if the PPS LLR is completed/reported *after* the start of the challenge? Or will the challenge WUs be served only after the start? Here are some of the scenarios I'm trying to avoid:
- I change properties now. After my PPS Siege current queue ends (which is expected to happen before challenge starts), Boinc sends me like 12h (or much more) worth of PPS LLR that will not be counted towards challenge because I downloaded them prior to start.
- I wait until challenge starts to change properties for new tasks. But, by then I'll have like 12h (or much more) worth of non-challenge WUs in my queue, which I don't want to neither abort nor expire.
So, what is the best approach?
Also, my workflow may be relevant:
I currently run PrimeGrid (among several other projects) 24/7 on Linux, 64 bits, with an AMD HD7770 GPU. The *only* task I ever got was "PPS (Sieve)" on my GPU (~1h:30m each). Unfortunately I also have to babysit the PC for PrimeGrid: it is the *only* project and task that makes my PC/mouse/desktop very laggy and choppy, So keep PG suspended when I'm using the PC (GPU then runs MilkyWay or Einstein), and I resume PG (while suspending all other projects) when I leave home or go to sleep. I hope LLR Sieve behaves differently.
So, for this scenario, how to fully maximize my participation in the challenge?
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
- I change properties now. After my PPS Siege current queue ends (which is expected to happen before challenge starts), Boinc sends me like 12h (or much more) worth of PPS LLR that will not be counted towards challenge because I downloaded them prior to start.
That is correct -- work you download BEFORE the challenge started will not count.
- I wait until challenge starts to change properties for new tasks. But, by then I'll have like 12h (or much more) worth of non-challenge WUs in my queue, which I don't want to neither abort nor expire.
If you do this, your computer will crunch 12 hours of non-challenge work prior to downloading tasks that WILL count for the challenge.
So, what is the best approach?
Set your cache to 0 instead of 12 hours and let your computer run dry (set "no new work".) Then once the challenge starts, hit the button to enable new work and then hit "Update". That's what most people do -- and why there's such a huge load on the server at the start of the challenge. This challenge is a little easier than most because it starts on a Sunday during non-sleeping hours (except for the U.S. west coast), so more people will be near their computers to mash on the "Update" button at exactly 11:02:00 UTC.
If you can not be at the computer at the start, your best bet is to set it to run PPP-LLR now, and set the cache to 0. With that setup, at worst you'll run 1 task per core that won't count for the challenge.
Also, you mentioned PPS-Sieve: Just to be clear, this challenge is for PPS-LLR, NOT PPS-Sieve.
____________
My lucky number is 75898524288+1 |
|
|
|
|
j.sheridan
PPS ( LLR ) - the best tasks ever for hungry CPUs
and GPUs ( PPS ( sieve ) I dont stop it, I have no problem with BoinC Server.
Everyone now thinks that makes the Boinc server problems.
The PC can also be spezifiziwer, depending from OS
If you mean GFN only,
GFN is runing / on my OS
Me 2.
I have PPS Sv and LLR running with no problems. I am beginning to wonder why me!!
If I do get the wrong ones I will just send them back (abort) as I would with anything else that was delivered from somewhere when I had not ordered it.
are your PCs setup with a short queue and finishing PPS in < 10 mins and PPS sieve in 15 mins?
|
|
|
|
|
|
And we're off!
Happy crunching and hunting for primes! :)
____________
PrimeGrid Challenge Overall standings --- Last update: From Pi to Paddy (2016)
|
|
|
LumiukkoVolunteer tester
Send message
Joined: 7 Jul 08
Posts: 165
ID: 25183
Credit: 870,450,997
RAC: 38,672
       
|
|
Standings not updating:
Participants | Teams
--
Lumiukko |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
Standings not updating:
Participants | Teams
--
Lumiukko
I just turned them on and they're now available.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Why is there two types of PPS_llr WUs, the PPS_llr run twice as long as the PPS_llr_Extended. Will more credit be given for the PPS_lls then the extended?
____________
 |
|
|
|
|
|
See Michael's post here, combined with this:
Scoring Information
... We will use the "prime score" method which is based on the n value (k*b^n-1) to score the challenge. The only difference is that the primary and double checker of a WU will receive the same score. |
|
|
|
|
|
And yes, the credit you receive is based on the on the length of the number you test. The longer the number, the longer the test, the more credit you get.
____________
PrimeGrid Challenge Overall standings --- Last update: From Pi to Paddy (2016)
|
|
|
|
|
|
Some one should check the person in first place. He is running Windows 8 and finishing tasks in 2 to 4 seconds. Looks like Win 8 is wreaking havoc again. |
|
|
|
|
|
I already noticed and notified Mike :) He has the magic touch that can actually do something about it, but he's also busy doing tons of other important stuff. (Like feeding the imps that give people work, as they seem be somewhat reluctant to help one person)
____________
PrimeGrid Challenge Overall standings --- Last update: From Pi to Paddy (2016)
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
Some one should check the person in first place. He is running Windows 8 and finishing tasks in 2 to 4 seconds. Looks like Win 8 is wreaking havoc again.
Don't worry about those. They won't count and will get automatically elliminated, even if we do nothing specific about it.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Hi,
it is my Win8 box that wasts all tasks. Does the wrapper have a problem with Win8? This computer is running well with other projects, e.g. POEM, but the llr-wrapper finishes after a few seconds. No error message, no exception. After a restart of windows, or a reinstall of PrimeGrid.....the same error?!
Some one should check the person in first place. He is running Windows 8 and finishing tasks in 2 to 4 seconds. Looks like Win 8 is wreaking havoc again.
Don't worry about those. They won't count and will get automatically elliminated, even if we do nothing specific about it.
____________
DeleteNull |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
Hi,
it is my Win8 box that wasts all tasks. Does the wrapper have a problem with Win8? This computer is running well with other projects, e.g. POEM, but the llr-wrapper finishes after a few seconds. No error message, no exception. After a restart of windows, or a reinstall of PrimeGrid.....the same error?!
Some one should check the person in first place. He is running Windows 8 and finishing tasks in 2 to 4 seconds. Looks like Win 8 is wreaking havoc again.
Don't worry about those. They won't count and will get automatically elliminated, even if we do nothing specific about it.
There's many Win8 systems running successfully.
There was a big discussion about Win8 few months ago, but I don't recall the details.
____________
My lucky number is 75898524288+1 |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
Hi,
it is my Win8 box that wasts all tasks. Does the wrapper have a problem with Win8? This computer is running well with other projects, e.g. POEM, but the llr-wrapper finishes after a few seconds. No error message, no exception. After a restart of windows, or a reinstall of PrimeGrid.....the same error?!
Some one should check the person in first place. He is running Windows 8 and finishing tasks in 2 to 4 seconds. Looks like Win 8 is wreaking havoc again.
Don't worry about those. They won't count and will get automatically elliminated, even if we do nothing specific about it.
It looks like this computer was successfully running PPS-llr tasks yesterday. Any idea what might have changed?
____________
My lucky number is 75898524288+1 |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
|
DeleteNull,
Here's what I can see so far:
Your computer was running these tasks just fine yesterday.
It's not a problem with the tasks themselves because they're working for almost everyone else. (There's only 7 computers that have returned faulty results like this.)
Your computer is reporting success, but it's not uploading a result file. It's probably not the cause, but could you check to make sure that your computer hasn't run out of disk space, and that BOINC hasn't used up all the space that's allocated to BOINC in your preferences?
And, it couldn't hurt to reboot the computer.
____________
My lucky number is 75898524288+1 |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
Some one should check the person in first place. He is running Windows 8 and finishing tasks in 2 to 4 seconds. Looks like Win 8 is wreaking havoc again.
The stats now automatically exclude bad results like that, so we don't have to wait for them to fail validation to be removed from the standings.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Hi,
>>>Your computer was running these tasks just fine yesterday.
Yes, yesterday i (re)attached this box to PrimeGrid, and all went well, only the challenge tasks today won't run.
>>>Your computer is reporting success, but it's not uploading a result file.
Yip, the wrapper finishes; but the llr-application hasn't started, so there is nothing to report.
>>>>but could you check to make sure that your computer hasn't run out of disk space
about one terabyte free space
>>>>and that BOINC hasn't used up all the space that's allocated to BOINC in your preferences?
every task has 2 gigabyte memory it may consume.
>>>>And, it couldn't hurt to reboot the computer.
Done two times, and two (extra) packages of tasks wasted. But don't worry: i am installing a linux in parallel. This will do the work ;-)
____________
DeleteNull |
|
|
|
|
|
My win 8 box was borking tasks so I rebooted into linux and all is fine now except my daughter won't be able to play SIMS for a few days. ;) |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
|
SCORING FORMULA CHANGE
It's been brought to my attention that there's a discrepancy in the challenge points being awarded to the shorter PPSE workunits and the longer PPS workunits.
Historically, for at least as long as I've been here, PrimeGrid has used the "prime score" formula for our LLR-based challenges.
This formula, although much, much better than using the old BOINC credit function, actually is not a very good formula to use during challenges. In particular, it favors larger tasks too much. (It's harder to find larger primes, and this formula is based on finding primes, not the time spent crunching the numbers.)
Therefore, I'm going to change the challenge formula to something more similar to the new credit function, so that it accurately represents the amount of work done. This change will be retroactive to the beginning of the challenge.
The raw number of challenge points everyone has will likely go up or down, en masse, once the change is made, but the relative positions shouldn't change much.
I don't like to change the rules in the middle of the game, but this is necessary to keep the challenge fair and have everyone playing on a level playing field.
Thank you in advance for your understanding.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Michael,
I'm not contesting your decision. However, on my main host (AVX intel cpu), short tasks take ~333 seconds. long tasks take ~960 seconds. The first were scoring ~20 credits and the latter ~58, which seemed proportional to the time spent crunching.
Assuming that the new formula will raise (proportionally) the credit to the short tasks I'm not sure how it will more "accurately represents the amount of work done".
____________
676754^262144+1 is prime |
|
|
|
|
The first were scoring ~20 credits and the latter ~58, which seemed proportional to the time spent crunching.
This is BOINC credit, for challenge used another score (~35 and ~171). |
|
|
|
|
|
Usucapio Libertatis,
It's been brought to my attention that there's a discrepancy in the challenge points being awarded to the shorter PPSE workunits and the longer PPS workunits.
short tasks = ~ 35 challenge points,
long tasks = ~ 171 challenge point
171 / 35 = 4.9
58 / 20 = 2.9
960 / 333 = 2.88
____________
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
Michael,
I'm not contesting your decision. However, on my main host (AVX intel cpu), short tasks take ~333 seconds. long tasks take ~960 seconds. The first were scoring ~20 credits and the latter ~58, which seemed proportional to the time spent crunching.
Assuming that the new formula will raise (proportionally) the credit to the short tasks I'm not sure how it will more "accurately represents the amount of work done".
You said "credits", and I'm assuming you indeed meant "credits" -- as opposed to challenge points. The problem is that the challenge points are wrong. Credits are correct. For LLR, they're not the same formula.
It's the challenge point formula that needs to be fixed. I'll probably make it the same as the credit formula, although I want to think about it a bit.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Thanks guys.
My mistake :(. I assumed llr credit (new formula) was being used for challenge points.
My apologies for the inconvenience, mainly to Michael, which has more to worry about.
By the way, serve seems to be working great, so congratulations.
____________
676754^262144+1 is prime |
|
|
|
|
By the way, serve seems to be working great, so congratulations.
hear hear. thanks admins. |
|
|
|
|
|
Some of my PPS LLR tasks are validated without a wingman. I thought LLR projects always needed two results to eliminate hardware and software errors? |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
Some of my PPS LLR tasks are validated without a wingman. I thought LLR projects always needed two results to eliminate hardware and software errors?
This is normal. Some smaller LLR projects permit "reliable" host computers to do the computation without a double checker.
____________
My lucky number is 75898524288+1 |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
|
The new scoring formula is now in place. The exact formula, for anyone who is interested, is:
challenge points = ((log10(k) + n * log10(b)) / 10000) ^ 2 / 20
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Is it me or it's *really* hard to get tasks from server?
It only handled me PPS LLR WUs in 2 occasions: like 10+ hours, a small batch of ~4 WUs, and more than 6+ hours ago a batch of ~8 WU. Not a single WU since then. Log says:
18-Mar-2013 03:09:18 [PrimeGrid] Sending scheduler request: Requested by user.
18-Mar-2013 03:09:18 [PrimeGrid] Requesting new tasks for CPU and ATI
18-Mar-2013 03:09:25 [PrimeGrid] Scheduler request completed: got 0 new tasks
18-Mar-2013 03:09:25 [PrimeGrid] No work sent
18-Mar-2013 03:09:25 [PrimeGrid] No work available for the applications you have selected. Please check your project preferences on the web site.
PPS LLR is the only application selected in my preferences
I am getting really upset of the immense amount of babysitting this challenge is requiring: every 2 hours or so I check Boinc Manager and instead of running PrimeGrid it is downloading and crunching for other minor priority projects. I have to suspend/abort their tasks, set to no new task, Update PrimeGrid, in hope it will get any work, then resume other projects.
Any directions on how to optmize and automate my participation? |
|
|
|
|
Is it me or it's *really* hard to get tasks from server?
It only handled me PPS LLR WUs in 2 occasions: like 10+ hours, a small batch of ~4 WUs, and more than 6+ hours ago a batch of ~8 WU. Not a single WU since then. Log says:
18-Mar-2013 03:09:18 [PrimeGrid] Sending scheduler request: Requested by user.
18-Mar-2013 03:09:18 [PrimeGrid] Requesting new tasks for CPU and ATI
18-Mar-2013 03:09:25 [PrimeGrid] Scheduler request completed: got 0 new tasks
18-Mar-2013 03:09:25 [PrimeGrid] No work sent
18-Mar-2013 03:09:25 [PrimeGrid] No work available for the applications you have selected. Please check your project preferences on the web site.
PPS LLR is the only application selected in my preferences
I am getting really upset of the immense amount of babysitting this challenge is requiring: every 2 hours or so I check Boinc Manager and instead of running PrimeGrid it is downloading and crunching for other minor priority projects. I have to suspend/abort their tasks, set to no new task, Update PrimeGrid, in hope it will get any work, then resume other projects.
Any directions on how to optmize and automate my participation?
BOINC will check often for GPU preferences first, as you have no GPU tasks selected but ATI is enabled you will have no work available. As BOINC is not to hot on handling CPU/GPU systems that is all it does.
Try deselecting GPU and then updating.
Also check the work computing preferences/ additional work in the Boinc Manager is set higher than usual for the challenge.
____________
Member team AUSTRALIA
My lucky number is 9291*2^1085585+1 |
|
|
|
|
I am getting really upset of the immense amount of babysitting this challenge is requiring: every 2 hours or so I check Boinc Manager and instead of running PrimeGrid it is downloading and crunching for other minor priority projects. I have to suspend/abort their tasks, set to no new task, Update PrimeGrid, in hope it will get any work, then resume other projects.
Boinc doesn't know you're participating in a challenge so if you don't set your other projects to "no new work" then it will keep fetching work for them if it thinks you need it.
Also, switch off the GPU in your primegrid preferences if you don't want GPU work (that's what the no work available message is about). |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
Is it me or it's *really* hard to get tasks from server?
Lots of different problems in your situation. Hopefully I'll touch on them all:
As for BOINC grabbing work from other projects instead of PrimeGrid, that's exactly what it's supposed to be doing. It's working as designed.
You may not agree with the way it's designed (I don't either) but it's not broken. Trying to get BOINC to download from multiple projects the way YOU want it to is an exercise in frustration. I threw in the white flag many years ago and only run a single project on each computer now. Otherwise, you really need to micro-manage it to get it to download what you want. It's not worth it.
As for the "no work" messages, that's complicated because there's a lot of causes. The one thing that's NOT happening is that the server isn't actually out of work. We've got new hardware and have made significant improvements to the database configuration, and the server is having no trouble keeping up with the workload. As far as I know, there's been no instances of the server not responding, or it not providing work because its buffer was depleted.
That's the good news.
The bad news is that with "server overload" out of the picture, we're seeing some real bugs crop up. In the past they were somewhat masked by the serious capacity problems, but they're obvious now.
Our version of the server is very old software, and it had some known (and unknown) bugs in it. I'm working on upgrading that software, but that won't happen until after the challenge. (The latest Boinc server software also has bugs in it, so I need to fix and test that before we can use it.)
One thing I discovered yesterday is that some users were being affected by a really bizarre server bug. Fortunately, there's a fairly easy fix for this. I'm going to apply that fix to your user ID, Let me know if it solves your problem.
(FYI -- the reason I don't do this globally for everyone is that only a few people seem to be affected, and the fix involves deleting the "trusted host" data for the affected computers. This means the computers are no longer trusted to do LLR tests without a double checker until they again return a lot of consecutive good results.)
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
We have found a treasure trove of primes! Currently there are 25 XGFN jobs pending on the server, which is quite a bit. We certainly are making short work
of PPS/PPSE during this challenge.
____________
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
We have found a treasure trove of primes! Currently there are 25 XGFN jobs pending on the server, which is quite a bit. We certainly are making short work
of PPS/PPSE during this challenge.
We're certainly finding a lot of primes during this challenge. Over 50 so far, and we're not even halfway through the challenge yet.
1.24 million tasks have been sent out, and 1.08 million tasks have been returned so far.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Are primes only displayed on a user's prime list after XGFN testing has been completed on the server?
____________
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
Are primes only displayed on a user's prime list after XGFN testing has been completed on the server?
To be honest, I'm not totally sure. There's a LOT of moving parts here, and I'm pretty sure that if you put the brains of Lennart, Jim, and myself into a blender, the resulting mixture would know all the details. :)
Therefore, I may have some of the details wrong, but this is how it works, more or less:
- When the second LLR result comes back and both results say the number is prime, the validator set a flag, adds 1 to the prime counts of both users, and add the prime into the primes database table.
- Every few minutes, a job runs that scans the LLR results looking for new primes, and adds them into the queue to be checked for XGFNs.
- When a CPU core is free (I'm currently running theXGFN checks on 3 cores although it's normally running on 2 cores), the prime is checked for XGFN factors.
- Once that check is done, Lennart uses a special interface to report the prime to the top 5000 list if the user wants us to automatically report primes on his behalf. A big delay factor is that often the user didn't fill in the "real name" part, or put in a bogus name or nickname. That website only accepts real names. If the user doesn't have automatic reporting enabled, we send them an email. They then have 19 days to respond, or else we give the prime to the double checker.
- Once the prime is submitted, the prime is verified by the top 5000 website servers, and once that's done, the prime is displayed on the top 5000 site.
- Periodically, PrimeGrid's server downloads the top 5000 primes from the top 5000 website, and once the prime is seen on the top 5000 site, the prime can be displayed on our site.
SGS primes do a different type of check (checking for a Sophie Germain pair).
The process for GFN is somewhat different and not as automated.
____________
My lucky number is 75898524288+1
|
|
|
|
|
|
Well, the server is handling the extra load from the challenge and the extra core running XGFN tests quite nicely.
____________
|
|
|
|
|
|
An interesting point about the time of the actual equinox. It occurs today, 20 March 2013, at 11:02 UTC (i.e. about 30 minutes from the time of writing this).
This means that 3 days (i.e. 60%) of the challenge would have taken place prior to the equinox and 2 days (40%) after.
____________
Warped
  |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3171
ID: 130544
Credit: 2,233,444,322
RAC: 550,199
|
|
Good point...
You either have it as 4-days, 6 days, or have the start-point 12 hours different. |
|
|
|
|
|
Not god points for this Tasks |
|
|
|
|
Not god points for this Tasks
The points are way better than they were before the new credit system was implemented for LLR tasks.
____________
|
|
|
|
|
Not god points for this Tasks
The points are way better than they were before the new credit system was implemented for LLR tasks.
... and at least you don't have to wait too long for the wingmen.
____________
Member team AUSTRALIA
My lucky number is 9291*2^1085585+1 |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3171
ID: 130544
Credit: 2,233,444,322
RAC: 550,199
|
|
I buggered up; had GFNs on pause & not enough cache & wasted all effing night not doing anything. Won't get that back. |
|
|
|
|
|
The first victory for Czech National Team, thank you all. |
|
|
|
|
|
Congratulations to the Czech National Team and to [DPC]x-RaY99_the_one_man_team.
Congratulations to the amazing 2545 users and 365 teams who participated on the challenge. If this is not a record, it should be very close.
Last, but not the least, congratulations to the organizers: work kept flowing flawlessly for five days and all of us that have been around for a while know how hard that is to accomplish in challenges with small tasks such as PPS or PPS_extended. Great work guys.
____________
676754^262144+1 is prime |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
|
At the Conclusion of the Challenge
We kindly ask users "moving on" to ABORT their WU's instead of DETACHING, RESETTING, or PAUSING.
ABORTING WU's allows them to be recycled immediately; thus a much faster "clean up" to the end of an LLR Challenge. DETACHING, RESETTING, and PAUSING WU's causes them to remain in limbo until they EXPIRE. Therefore, we must wait until WU's expire to send them out to be completed.
Please consider either completing what's in the queue or ABORTING them. Thank you. :)
On PrimeGrid's front page, there will be a notification ("*** Challenge (PPS-LLR) cleanup work IS currently available! ***") near the availble work numbers when there's currently tasks available that will help with the cleanup.
Some interesting statistics:
Total tasks completed during the challenge: 2,262,786
--PPSE: 1,345,453
--PPS: 917,333
Total PPS & PPSE numbers crunched: 2,063,210
--PPSE: 1,237,353
--PPS: 825,857
Primes found during the challenge: 138 (*)
--PPSE: 100
--PPS: 38
(*) The number of primes is the number of validated primes where either the prime finder or the double checker processed the prime during the challenge. Therefore, there's probably more primes that were found during the challenge, but we don't see them yet because they're not double checked. This number will likely increase as pending results eventually get validated.
____________
My lucky number is 75898524288+1
|
|
|
|
|
Congratulations to the Czech National Team and to [DPC]x-RaY99_the_one_man_team.
Congratulations to the amazing 2545 users and 365 teams who participated on the challenge. If this is not a record, it should be very close.
Last, but not the least, congratulations to the organizers: work kept flowing flawlessly for five days and all of us that have been around for a while know how hard that is to accomplish in challenges with small tasks such as PPS or PPS_extended. Great work guys.
A big hearty amen and +1 !
____________
@AggieThePew
|
|
|
|
|
|
Na zdraví, Češi |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
|
Cleanup Status:
March 22nd: 11477 tasks outstanding; 2141 in individual (73) scoring positions; 51 in team (5) scoring positions.
March 23rd: 8250 tasks outstanding; 1217 in individual (56) scoring positions; 37 in team (5) scoring positions.
March 24th: 4666 tasks outstanding; 484 in individual (38) scoring positions; 12 in team (2) scoring positions.
____________
My lucky number is 75898524288+1 |
|
|
|
|
|
Fine job Czech National Team and congratulations for your first place! :)
Also special thanks to the organization team for their quick fix at the challenge formula to keep the challenge fair. :)
The overall statistics looks imressive - well done to all participants! |
|
|
|
|
Primes found during the challenge: 138 (*)
--PPSE: 100
--PPS: 38
(*) The number of primes is the number of validated primes where either the prime finder or the double checker processed the prime during the challenge. Therefore, there's probably more primes that were found during the challenge, but we don't see them yet because they're not double checked. This number will likely increase as pending results eventually get validated.
http://www.primegrid.com/workunit.php?wuid=329020880
2195*2^1095881+1 is prime!!!
I guess I am the last primefinder of the challenge:
22 Mar 2013 | 10:39:58 UTC
____________
|
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
|
Cleanup Status:
March 22nd: 11477 tasks outstanding; 2141 in individual (73) scoring positions; 51 in team (5) scoring positions.
March 23rd: 8250 tasks outstanding; 1217 in individual (56) scoring positions; 37 in team (5) scoring positions.
March 24th: 4666 tasks outstanding; 484 in individual (38) scoring positions; 12 in team (2) scoring positions.
March 25th: 611 tasks outstanding; 11 in individual (5) scoring positions; 1 in team (1) scoring positions.
____________
My lucky number is 75898524288+1 |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
|
Challenge stats are final!
____________
My lucky number is 75898524288+1 |
|
|
Dave
Send message
Joined: 13 Feb 12
Posts: 3171
ID: 130544
Credit: 2,233,444,322
RAC: 550,199
|
|
I went onto GFN at the right time then... |
|
|
Michael GoetzVolunteer moderator
Project administrator
Send message
Joined: 21 Jan 10
Posts: 13955
ID: 53948
Credit: 392,700,090
RAC: 171,419
|
Primes found during the challenge: 138 (*)
--PPSE: 100
--PPS: 38
(*) The number of primes is the number of validated primes where either the prime finder or the double checker processed the prime during the challenge. Therefore, there's probably more primes that were found during the challenge, but we don't see them yet because they're not double checked. This number will likely increase as pending results eventually get validated.
We ended up barely breaking 150 primes for the challenge!
Final update on the number of primes found:
Total primes: 152
--PPSE: 109
--PPS: 43
____________
My lucky number is 75898524288+1 |
|
|
